The Porgera gold deposit in Papua New Guinea is a world-class example
of an alkalic-type epithermal gold system (stage II), which overprints
a precursor stage of magmatic-hydrothermal gold mineralization (stage
I). Gas and ion chromatographic analyses of fluid inclusions containe
d in vein minerals from both mineralization stages have been carried o
ut in order to constrain the compositions of the fluids involved in, a
nd the processes attending, ore deposition. These data indicate the pr
esence of three end-member liquids, the most dilute of which was prese
nt throughout the mineralization history and is interpreted to represe
nt evolved groundwater of meteoric origin. Its composition is estimate
d to have been approximately 500 mM Na+, 10 mM K+, 5 mM Li+, 250 mM Cl
-, 0.15 mM Br-, and 0.01 mM I-, plus significant concentrations of dis
solved gases. More saline liquids were also present during the two mai
n stages of ore formation, and although their compositions differ, bot
h are interpreted to have been derived at least in part from magmatic
fluids, and to have been the media by which gold was introduced into t
he system. Stage I minerals contain fluid inclusions which decrease in
salinity towards this dilute end-member composition through the vein
paragenesis, reflecting progressive dilution at depth of the magmatic
fluid source by groundwaters. Ore deposition is thought to have been c
aused largely by simple cooling and/or wallrock reactions, although li
mited in situ fluid mixing may also have occurred. The most saline flu
ids, present in early quartz and pyrite, contain at least 810 mM Na+,
530 mM Ca2+, 130 mM K+, 12 mM Li+, 87 mM SO42- 960 mM Cl-, 1.1 mM Br-,
and 0.05 mM I-, plus significant but variable concentrations of disso
lved gases. Fluid inclusions from stage II hydraulic breccia veins rev
eal the presence of two distinct liquids with contrasting salinities,
which were present at different times during vein formation. A higher
salinity liquid appears to have predominated during mineralization, wh
ereas lower salinity groundwaters filled the structures during interve
ning periods. The ore-forming fluid may have been forcibly injected in
to the veins from depth during fracturing and depressurization events,
displacing the resident groundwaters in the process. The original com
position of this fluid is estimated to have been at least 1770 mM Na+,
59 mM K+, 180 mM Li+, 210 mM SO42-, 680 mM Cl-, 1.4 mM Br-, and 0.09
mM I-, plus 1.5 mol% CO2, 0.19 mol% CH4, and 0.04 mol% N-2. Gas chroma
tographic analyses of fluid inclusions from stage II samples show a de
crease in total gas content between early unmineralized veins and post
-mineralization vuggy quartz (suitable samples could not be obtained f
rom the ore stage itself). Post-mineralization samples plot along an e
xperimental gas-saturation curve in the CO2-CH4-H2O-NaCl system, obtai
ned at conditions similar to those attending stage II ore deposition a
t Porgera (200-300 bar, similar to 165 degrees C). These results are i
nterpreted to indicate a period of depressurization-induced phase sepa
ration during hydraulic fracturing, which resulted in rich ore deposit
ion. Volatile gases such as CH4 and N-2, in addition to CO2 in solutio
n, are shown to have a significant negative effect on total gas solubi
lity. This effect may be of critical importance in lowering the temper
ature and increasing the depth (pressure) at which phase separation ca
n occur in epithermal systems.